If you do not fully understand the audience, planning a\/b test activities will become a guessing game. Therefore, it is important to investigate first. The best place to start is Google Analytics. With Google Analytics, you can learn more about who visitors are, from which country, age and gender, screen resolution and browser used on specific devices, how to reach the website, and in which language. They say, even what they care about. All this information is available, so there are many potential customers to explore and understand. These potential customers help to plan a\/b testing activities for users more effectively. The
One of the most recent posts was on WordPress, testing the 11 best tools for a\/b, what to use as the details of why \/b testing. In today’s post, we will learn about some metrics that can be used and used when planning split test activities. The key point of using Google Analytics’ reason segmentation test is not knowing. Your task is to make educated guesses about what the most useful content \/ design is, but you can’t actually go beyond that. With split testing, you can test assumptions and make decisions based on actual data rather than speculation. The
That is, you want to test as efficiently as possible. Focus on the most promising candidate. Deciding what the candidate is is also the subject of discussion. Here, Google Analytics may help. You can use the various metrics provided (technical competence, age, interest, user flow, etc.) to narrow potential candidates to manageable ranges. To provide meaningful data, the test must be important. If you don’t have millions of visitors a day, you can’t do many segmentation tests. Although I don’t want to talk about math here, there are a lot of reports on the importance of split testing and other a\/b test errors on conversionxl. If you plan to do a split test, be sure to read it! The
The first is the indicators we can use. We can use all Google Analytics measurement projects. All the information we have can potentially narrow the scope of design or content decisions. Let’s look at some of the points used to make decisions in the past. But this is not a complete list. Depending on the focus area, other measurement items may be more useful. The idea behind all this is more important. If you master it, nothing can stop you! Age, gender age and gender are very important factors in design and content. Compared with middle-aged women, dialogue with young men requires a variety of behavioral guidance texts. The
These are usually the first two measurement items that I started looking at in Google Analytics. Leads > can be found in demographics. The number of visitors’ age distribution is obtained from the website I manage. In order to protect my personal information, I will not specify a name here. Obviously, the website is aimed at the young side of the age spectrum. The age group of 25~34 years old is the most prominent, but the age group of 18~24 years old follows closely. This kind of information can allow web designers to explore newer and younger designs, but it is not very different from what people are familiar with today. The
Similarly, data affects duplicate lights. On such sites, you can use move now as a CTA button to advise users to launch. If we target a much older age group, we may not deviate from the very clear \
Your site will introduce default language groups, but you may encounter some examples of important language groups. Canada would be a good example of having two official languages. Language and location data can be found in target > region, as follows: The
Male: among the top 6 countries, 4 are native speakers of English, but 7.65% of visitors are from India. Perhaps by analyzing cultural differences, we can come up with special images that are more likely to attract the attention of Indian friends. The session analysis objectives > behaviors section provides information about the session. You can view new and re visitors, the number of sessions per visitor, and participation. Let’s take a look at the frequency view, which highlights the key points that can be used in the male segmentation test work. Male session frequency this view shows 171886 single sessions out of a total of 240657 sessions. To understand this, you must know what a session is. Google Analytics introduces session definitions. In short, a session is usually a group of activities that a user performs within a given period of time (30 minutes). The
So our visitors tend not to use the website at least once a month (the total data I see is one month). This will prompt you to navigate to tools that can redirect visitors back to the site. This can be as simple as a\/b test release frequency, but as fine as splitting and testing various design and behavior guides on the press release registration form. In the technical part of Google Analytics, the most important indicator used when planning a\/b tests is screen resolution. You can find this information by navigating to goals > Technology > browser and operating system. There is a default measurement standard selector above the data table, which can be used to switch to the screen resolution view or other measurement items. The
A conclusion drawn from the data above the screen resolution information is that better visual objects should be segmented and tested on small mobile devices. Devices with a width of 300-400 pixels account for 13% of the traffic, but the average navigation time is only 25 seconds. This can be said to be an inevitable attribute of small machines, but I can point out different implications. Visitors to 360x640px devices spend an average of 1 minute 06 minutes viewing content, and 375x667px devices spend 41 seconds. This may be accidental or a property of the device. Although more analysis is really needed here, it can indicate the right (or wrong) direction of the test. The
Either party can split and test the partial deformation of mobile device content display, which will help guide more mobile views, or make the existing mobile viewer happier, thus increasing the revenue of the site. Through in-depth observation, the deviation rate of this resolution can be significantly increased to 99.98. Acquisition shows where visitors come from. Categories include natural search, direct, referral, email, social, and others. Regardless of the appearance of specific data, you can make split test decisions on this basis. Sample sites include: The recommendation shows that more than 86% of the website traffic comes from natural search. You can do this by detecting search content and building facilities that provide similar content. Users can continue to search on the Google site without returning to Google. The
If there are a lot of recommended traffic, \
Although I don’t tell you exactly what to do to improve, I provide reliable statistics to help you decide where to go. The next step in this process is the split test, which can tell you the most useful exact deformation according to the instructions. If you want to learn more about a\/b split testing, don’t forget to check out the 11 best tools WordPress uses for a\/b split testing mentioned earlier. Have you used Google Analytics to plan a\/b split test activities? What types of segmentation tests are performed on our website? Please tell us the following comments. Labels: Test Environment
